Stop reusing the Prism result for RuboCop for now #3672
Merged
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.

Motivation
Related to #3657 and #3665. This PR "fixes" the issues by not reusing the Prism result anymore, but the real fix is on the RuboCop side.
Reusing the Prism parse result for RuboCop is a massive performance win as we avoid double-parsing the same document - both for computing diagnostics on every keypress and for formatting.
Unfortunately, we are receiving numerous reports of RuboCop crashing when the Prism result is reused, so I propose we turn this off while the RuboCop team investigates why the cops fail in this mode.
It might be easier to add some sort of experimental setting, so that people can find bugs. It seems like the problems are not related to running RuboCop in the CLI, where it sees all files in their final state, but in LSP mode when the user is in the middle of typing.